Micro Machining Milling Machine Option

[implmex]

Hi Kingbob:
I appreciate your comments; here's my take on it all:
I had the great pleasure of building a shop for myself while I was still in my prime and I got to make some really amazing things for some really great people.
I indulged my acquisitive lust for many years, unwrapping my new toys and playing with them with pride and enthusiasm.

Now I find myself at 66, with a neurodegenerative condition (progressive MS) and a slowly declining ability to build the stuff that used to give me so much pleasure.
In addition my son is a computer geek with zero interest in what I do, and has recently pleaded with me not to make my shop full of obsolescent toys his problem when I'm no longer able to drag my sorry ass in here.

So much as I appreciate the sentiment, and much as I agree with your point in principle, sadly new machine shop toys are not in my plans anymore.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com

 

[thesidetalker]

Tolerances I usually work with are +/-.0002" (this is usually only one or two dimensions on a print and are usually less than .1875").

Are these feature sizes? or positional?

If you have things needing to be lined up this close on different sides of the part, Haas shouldn't be in the list. It might hold the size of a feature, but once you start rotating things, just forget it.

Don't forget to consider climate control as well. My shop is not and has wild temperature swings from morning to afternoon. Currently running a job on my Matsuura, which needs a feature lined up from one side to the other within 12 microns, meaning rotary center parameter needs to be within 6 microns. It is a pain in the ass as it drifts quite a bit with temperature and requires frequent calibration to keep it updated.

I wouldn't even bother trying to do this on my UMC750, even if I could probe the first feature from the second side. Even then its a gamble. Then what happens when the probe gets some dust and dried up coolant on it? Throw your position way off.

The CM-1 might be well within your budget, but you could possibly waste many hours and thousands of dollars of time fighting it... as mdk says:

There's no $3000 Haas payment that can be made to be competitive with what a premium controller/builder can do that would require $6k-8k payment.
Price is what you pay, values is what you get.
Paying a guy $5000 per month to dic around with fixture offsets due to 5 axis inaccuracies is certainly part of the puzzle that doesn't show up in the 'build a quote's.

 

[Conrad Hoffman]

Some years ago I thought about starting a high accuracy micromachining shop. The cost of the machines seemed high, but manageable. What killed it? The cost of all the equipment needed to verify that the parts actually met the prints, plus ongoing certification of everything. Every time you move the decimal place, the costs go up by two or so.

 

[Rcgiovannani]

Are these feature sizes? or positional?

If you have things needing to be lined up this close on different sides of the part, Haas shouldn't be in the list. It might hold the size of a feature, but once you start rotating things, just forget it.

Don't forget to consider climate control as well. My shop is not and has wild temperature swings from morning to afternoon. Currently running a job on my Matsuura, which needs a feature lined up from one side to the other within 12 microns, meaning rotary center parameter needs to be within 6 microns. It is a pain in the ass as it drifts quite a bit with temperature and requires frequent calibration to keep it updated.

I wouldn't even bother trying to do this on my UMC750, even if I could probe the first feature from the second side. Even then its a gamble. Then what happens when the probe gets some dust and dried up coolant on it? Throw your position way off.

The CM-1 might be well within your budget, but you could possibly waste many hours and thousands of dollars of time fighting it... as mdk says:

I am talking about features of size. Rarely do I have something that is less than 0.1mm positional tolerance.

I forgot to mention that I was also looking at the Okuma Genos M460-VE and M560-V. I would get a Koma 5 axis trunnion with one of those. I love Okuma machines - I don't own any but have used them before I owned my own shop.

I kind of gave up on that idea since the machines only go up to 15k RPM. Maybe it's better to have good thermal stability and accuracy / repeatability than it is to have more RPM's? I was looking into Coolspeed (saw it in an Edge Precision video on YouTube). Maybe a more accurate, bigger machine with spindle speeders is the answer?

These are the questions I keep coming back to... Not sure if there is a right answer.

 

[implmex]

Hi Rcgiovannani:
I get away with 6K rpm...15K would be heaven for me but for cutters under 1/2mm, 25K would be even happier.
All depends on your mix of work...as soon as you depend on making time on your parts the speed pays off very quickly, and the mass of a hunky machine like a Genos begins to get in your way because you are never going to make it work hard with a 0.020" cutter so you don't need the rigidity.

If you have too much mass to push around, you are wasting poop just to keep the spindle turning and the table flying around.
When you've got some godawful massive lump being gnawed to shape , you need all that machine beef too, but if you don't, it's a waste.

Modern spindle speeders are nice, I admit, but they eat up Z axis stroke and they are so very very fragile.
One crunch and it's crunched...zero tolerance to misadventures.

I personally would probably opt for a nice Brother for the kind of parts you talked about.
If you need to process the occasional bigger part, that Genos will make you very happy too, even with its limitations for the micro world.
Both will make you nice parts, but if you do mostly micro, you're leaving money on the table with the Genos.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com

 

[Rcgiovannani]

Hi Rcgiovannani:
I get away with 6K rpm...15K would be heaven for me but for cutters under 1/2mm, 25K would be even happier.
All depends on your mix of work...as soon as you depend on making time on your parts the speed pays off very quickly, and the mass of a hunky machine like a Genos begins to get in your way because you are never going to make it work hard with a 0.020" cutter so you don't need the rigidity.

If you have too much mass to push around, you are wasting poop just to keep the spindle turning and the table flying around.
When you've got some godawful massive lump being gnawed to shape , you need all that machine beef too, but if you don't, it's a waste.

Modern spindle speeders are nice, I admit, but they eat up Z axis stroke and they are so very very fragile.
One crunch and it's crunched...zero tolerance to misadventures.

I personally would probably opt for a nice Brother for the kind of parts you talked about.
If you need to process the occasional bigger part, that Genos will make you very happy too, even with its limitations for the micro world.
Both will make you nice parts, but if you do mostly micro, you're leaving money on the table with the Genos.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com

Thank you, Marcus. I've got a call with Yamazen this afternoon to get a quote on a Brother S700X2. Hoping that leads to somewhere for me haha.

 

[mhajicek]

With tiny parts, your cycle time will be 95% determined by the spindle RPM. Better motion control and faster toolchanges could shave that last 5%. Is a 5% shorter cycle time worth 5x to 10x the machine price? Maybe once you're established and doing long runs, but not for your first machine. And if the tolerance is comfortably good enough, it's good enough. You don't get paid more for being .001mm from nominal instead of .010mm, as long as both are well within spec and the part looks good and works as intended.

 

[hi-fly-cnc]

With tiny parts, your cycle time will be 95% determined by the spindle RPM. Better motion control and faster toolchanges could shave that last 5%. Is a 5% shorter cycle time worth 5x to 10x the machine price?

Can I gently disagree here? And I say this, stating up front that I'm a beginner and almost certainly less experienced than yourself, I'm only making an observation.

I would agree if we were to imagine that we had already bought the best tools possible, optimised the program for the machine perfectly and were dominated by time in the cut. However, my limited experience suggests that variables such as machine accelerations, tool runout and to some extent tool quality (eg CTS vs non CTS) are able to make quite a significant difference on cut time?

Here's a post I put on my insta where I was running the same faceplate operation (0.5mm to 3mm tools) and the only difference was knocking the rapid knob on the Speedio from 5 to 4. Difference was 4:20mins to 5:10 mins cycle time. So not far off a minute slower or 25% slower, just by reducing the speeds of the rapids, ie slowing down *non cutting* time (you can see some pictures of the finished part in other posts - it's a faceplate for an electrical enclosure)

https://www.instagram.com/p/CiGqZmaDQsy/

Now, I don't claim that someone else couldn't create a far better program! However, just an anecdote: On this program, much of the cutting time is taken on the tiny LED holes that are 1.5mm across and something like 3mm in length. However, the Brother has several features which allowed me to drop the machining time by about a factor of 3.

- Tool change time is so low (sub 2 seconds) that I can do a 1.5mm slot op, then a 1mm rough, then a 0.5mm finish. Using 3 tools is a massive time save (over say 2 tools, or even using only the 0.5mm alone).

- Acceleration is >1G. This means you can hit 10,000mm/min in <1mm of distance. So for example a similar part has a 2.1mm wide slot for a SIM card. The adaptives into the corner of that 2.1mm slot (0.08") have a 10,000mm/min non engagement move set on them and quite often it's hitting that (or close to) on repositioning moves!

- Same bonkers acceleration on the Z (actually it's faster, >2G accel on Z...). So repositioning moves pop up and down into the cut at 10m/min on my 0.5mm tool, could go faster still... This often halves my op time vs if instead I had forgotten to change the tool defaults

- Repositioning across the table, rapids are 50k/min. Sometimes Fusion360 does stupid moves where its adaptive will jump to both extremes of the job. I have in the past invested time getting rid of this, only to find it saves sub second time on the job... Not saying it's good to have an inefficient op, but the time cost for this is not so high (the programmer time is not free)

I would add high quality collets or tool holding + CTS as having a big effect on cut time. However, it's true that those could be applied to any (decent) machine. I find I can push tools much harder than I ever expected now (and that the manufacturer recommended...)

I don't think the Brother machine necessarily counts as a "super accurate" micro milling machine. To be fair I suspect other brands might do this job better and more accurately? I certainly find the dimensional accuracy in X and Y is accurate beyond my measuring equipment/ability to check it. However, work holding variations mean that I am probably worse than 0.02mm in Z on a regular basis. Not important for my work, but worth noting if that were part of the brief

However, what DOES impress me is watching the tool change for the 0.5mm tool when on full rapids... The old tool flies up, you get the Brother tool change whir! Then that 0.5mm comes down in a blink of an eye at 50,000m/min towards the solid fixture . Fusion has it rapiding into empty space, to stop at z=0 !!! There is nowhere further for it to go!! Then it turns right and starts machining out the corners of the pocket... Every time it changes tool I'm always amazed that the tool is still there...! It impresses me anyway!

 

[implmex]

Hi hi-fly-cnc:
That was a great and informative post...I really enjoyed reading it.

However mhajicek makes a great point too, and I believe you're observing differences that come from the nature of the work each of you are taking on.

When your parts are mostly prismatic and lots of toolchanges are involved, and you do better taking the maximum cutter diameter the part geometry and the cut will allow, your strategy is the right one, and your observations of course are correct.
However, as soon as your parts involve surfacing or you need to accept tiny chiploads (like with a fragile cutter hanging out a long way in a difficult material), the spindle speed starts to dominate the cycle time because you have such a very long cut path, and obviously the faster you can turn it without burning up the cutter, the faster you can drive it without snapping it off.

So both are very nice to have.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com

 

[Rcgiovannani]

Hi guys,

Just a quick update. I ended up going with a Brother Speedio S700X2 with 27k spindle and a Lyndex Nikken 5AX-130SA-BR2 Rotary Index table. Got a pretty good deal from Yamazen since it was the week of IMTS and the machine is a 2020 demo machine. Should be delivered and setup in 3-4 weeks!

 

[implmex]

Hi Rcgiovannani:
Make me jealous why don'tcha
That is great...here's to your new machine.
I wish for many happy hours making really cool things to an amazing level of precision and just so damned lovely I will have tears of joy in my eyes.

Seriously...congratulations!

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com

 

[Rcgiovannani]

Hi Rcgiovannani:
Make me jealous why don'tcha
That is great...here's to your new machine.
I wish for many happy hours making really cool things to an amazing level of precision and just so damned lovely I will have tears of joy in my eyes.

Seriously...congratulations!

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com

Thanks, friend!

 

[LOTT]

Hi guys,

Just a quick update. I ended up going with a Brother Speedio S700X2 with 27k spindle and a Lyndex Nikken 5AX-130SA-BR2 Rotary Index table. Got a pretty good deal from Yamazen since it was the week of IMTS and the machine is a 2020 demo machine. Should be delivered and setup in 3-4 weeks!

Congrats, that should be an awesome machine for your parts; you get the best of both worlds with RPM and the positional speeds.

Just out of curiosity, did you compare that setup to a U500? I know you aren't making large parts that need the swing, I'm just wondering if the integrated trunnion would be nicer/more flexible. I also don't know what the price comparison is.

 

[Rcgiovannani]

Congrats, that should be an awesome machine for your parts; you get the best of both worlds with RPM and the positional speeds.

Just out of curiosity, did you compare that setup to a U500? I know you aren't making large parts that need the swing, I'm just wondering if the integrated trunnion would be nicer/more flexible. I also don't know what the price comparison is.

I'm not sure I know what a U500 is. Haas UMC500? I googled U500 and something called a Chester U500 popped up but I've never heard of that MTB.

I got an absolutely killer deal on the S700X2. Even our finance guy was like "whoa, that's a good deal on one of those machines" lol.

 

[hanermo]

Excellent machine !
Congrats.
Whats the ballpark monthly nut on one of these ?

 

[LOTT]

I'm not sure I know what a U500 is. Haas UMC500? I googled U500 and something called a Chester U500 popped up but I've never heard of that MTB.

I got an absolutely killer deal on the S700X2. Even our finance guy was like "whoa, that's a good deal on one of those machines" lol.

Speedio U500, just came out.

https://machinetool.global.brother/en-ap/u500xd1

I imagine it would make more sense for larger parts, for the micromachining you mentioned the S700 is probably perfect. Not meaning to distract from your thread, it's just new and I'm a Brother fan, so I'm curious about it.

Back on topic- if you're able to post the S700 delivery and first parts you make we'd love to see them.

 

[BROTHERFRANK]

Speedio U500, just came out.

https://machinetool.global.brother/en-ap/u500xd1

I imagine it would make more sense for larger parts, for the micromachining you mentioned the S700 is probably perfect. Not meaning to distract from your thread, it's just new and I'm a Brother fan, so I'm curious about it.

Back on topic- if you're able to post the S700 delivery and first parts you make we'd love to see them.

I think the S700X2 with 27K spindle and tilting rotary was the right call. Great set up and you still have room for vises on the table... Congrats OP. Can't wait to see your new machine day post and hear about your first parts.

 

[Rcgiovannani]

I think the S700X2 with 27K spindle and tilting rotary was the right call. Great set up and you still have room for vises on the table... Congrats OP. Can't wait to see your new machine day post and hear about your first parts.

I will for sure be posting my new machine day!

Already started laying out the table. I'm going to be going with 5th axis RockLock system. Probably won't be getting the dual station vice in the picture, will most likely just get a Kurt Vice adapter for the 96mm base (for now that's all my budget will allow for). But I will be getting the stuff shown on the rotary, plus an ER40 adapter for that as well.



It's been pretty overwhelming tooling up this machine since it's my first 5-axis and also my first 30 taper machine so I'm pretty much starting from scratch. I can't wait.

Was going to splurge on some RegoFix PowrGrip stuff (which I knew was expensive) and then I saw the actual price and almost fainted lol. I'll be doing Lyndex Nikken SK6 and SK10 and some High speed milling chucks for holders.

 

[BROTHERFRANK]

That looks like a great set up. You are well on your way. As far as tooling goes, the Nikken SK chucks are nice. I like the High Speed (smooth) nuts on them for high rpm. Keeps coolant whip down. Run Nikken label SK collets with them. Seem to work best. Keep tool holders as short as possible. As much as I like RegoFix products, I find the Powrgrip offerings for BT30 tend to have long gage length which negates some of the benefit. I like stubby end mill (side lock) adapters for roughing. I ran a 12mm side lock with a weldon flat rougher at 16k the other day, smooth as silk. MST shrink fit are nice if you need some reach with tight clearance issues.

 

[Rcgiovannani]

I was thinking I needed longer tools since I have the Tilting rotary. Was going to buy some of the "middle length" tools like BT30-SK6-90P and BT30-SK10-90P. Both with the high speed nuts. the picture has the SK6-90P. Clears the vice nicely.